1. Field of the Invention
The present invention relates to a rotary machine diagnostic system.
2. Description of the Related Art
In a case where an unexpected failure occurs with a rotary machine such as a motor (electric motor) or a generator assembled into a production facility, it is necessary to perform unplanned repair work or unplanned replacement work on the rotary machine such that the rate of operation of the production facility is decreased, or it is necessary to revise a production plan.
A deteriorated state may be ascertained by stopping the motor at a suitable timing and performing a diagnosis for the motor offline so as to forestall an unexpected failure, and as a result, it is possible to prevent an unexpected failure to some extent. However, it is necessary to stop the rotary machine for offline diagnosis, thereby causing a decrease in the rate of operation of the production facility.
From the above background, there are increasing demands for online diagnosis by which a rotary machine can be diagnosed during operation. Online diagnosis also allows a system to detect symptoms of the rotary machine which are obvious only during rotation.
For example, one of causes of an unexpected failure is insulation damage to a rotary machine. An example of the related art regarding insulation damage to a rotary machine is disclosed in Non Patent Literature (NPL) 1: P. Zhang, K. Younsi and P. Neti: “A Novel Online Stator Ground-Wall Insulation Monitoring Scheme for Inverter-Fed AC Motors”, IEEE Transactions on Industry Applications Vol. 51 No. 3 (2015) pp. 2201-2207.
As a diagnosis technique for a rotary machine, NPL 1 discloses a method of calculating a parameter of a leakage current path equivalent circuit from a current equivalent to a leakage current and the phase difference and the amplitude ratio of a voltage applied to a motor via fast Fourier transform (FFT).
However, the method disclosed in NPL 1 has the following problems.
According to the FFT technology disclosed in NPL 1, current and voltage waveforms are required to be in a static state in order to allow accurate extraction of equivalent circuit parameters. In contrast, in a motor that is actually being used, a rotational speed may change or a load connected to the motor may change, and thus, the current or voltage of the motor may change. In a case where a signal having a frequency component used to extract a capacitance is small compared to the noise, there occurs a problem that a capacitance extraction error becomes increased. If the capacitance extraction error is large, in a case where a voltage applied to the rotary machine changes transiently due to change in a load or the like driven by the motor, there occurs a problem that it becomes difficult to accurately calculate a parameter of a leakage current path equivalent circuit.